In the field of image processing, a similar image recognition technology becomes more and more mature. At present, a common image similarity recognition method includes a color distribution histogram CF algorithm and a perceptual hash (phash) algorithm. The CF algorithm has a very high error recognition probability, that is, it is very likely to mistakenly determine an image dissimilar to a reference image to be similar. The phash algorithm may ensure recognition accuracy to a certain extent, but is likely to miss some pictures with relatively high similarities. As can be seen, there is an urgent need of a solution capable of preventing pictures with high similarities from being missed and also ensuring or improving recognition accuracy.